6G Orchestration and Automation: A System Software View

Orchestration and automation (O&A) are the machinery for optimally configuring a service provider’s assets in view of its chosen business models. At the moment, business models are diversifying, especially at the edge of the network. The emergence of neutral hosts and the offloading of enterprise and end user workloads to edge clouds are but two examples of the ongoing developments. Consequently, O&A needs to cater to a variety of business models and value chains. Further, 6G needs to support interactions between stakeholders participating in service provisioning, which is more dynamic than the current paradigm based on relatively static service level agreements (SLAs) and contracts. Given the expected proliferation of business models and use cases, interoperability will be an important factor for the 6G era. 6G multi-access networks will often be built on an existing 5G network, allowing for an easy introduction of 6G capabilities for service providers as well as the ability to monetize resources right from the launch, an important factor in the success of 6G. 3GPP standards provide a global backbone for interoperability in 6G, supported by ancillary standards, best current practices (BCPs) and de facto standards. Automation is a key requirement for 6G, which, because it is expected to natively incorporate artificial intelligence and machine learning (AI/ML), the entire architecture has been designed with AI/ML support in mind. This also includes data architectures supporting training and inference. The impact of AI/ML nativity on O&A is twofold: on the one hand, O&A itself will have advanced AI/ML capabilities; on the other, it will need to manage AI/ML features for the rest of the 6G system.